Geology topics

J. Dimasaka

Publications and source records attributed to J. Dimasaka.

3 recordsLinked to original sources

USGS2023

Causality-informed Bayesian inference for rapid seismic ground failure and building damage estimation

Rapid and accurate estimates of seismic ground failure and building damage are beneficial to efficient emergency response and post-earthquake recovery. Traditional approaches, such as physical and geospatial models, have poor accuracy and resolution due to large uncertainties and the limited availability of informing geospatial layers. The introduction of remote sensing techniques has shown potential in providing supplementary information for rapid hazard estimation by analyzing earthquake-induced correlation changes between pre- and post-event satellite images. However, the changes in satellite images are the result of overlapping ground failure, building damage, and environmental noise, making it challenging to categorize and estimate different seismic hazards and impacts directly from satellite images.Here we design a novel causality-informed Bayesian network that continuously updates seismic ground failure and building damage estimates from satellite images by modeling the physical interdependencies between geospatial features, ground failure, building footprints, building damage, and satellite images. The incorporation of physical interdependencies allows an effective fusion of physical models and rich but noisy information from remote sensing observations and reduces bias and uncertainties in estimations. Our experiments show that integrating satellite images through our Bayesian network improves the accuracy of seismic ground failure and building damage estimations.

Conference Paper↗

USGS2022

Integrated strategies for enhanced rapid earthquake shaking, ground failure, and impact estimation employing remotely sensed and ground truth constraints

Estimating earthquake impacts using physical or empirical models is challenging because the three components of loss estimation-shaking, exposure, and vulnerabilities-entail inherent uncertainties. Loss modeling in near-real-time adds additional uncertainties, yet expectations for actionable information with a reasonable level of confidence in the results are real. The modeling approaches described herein augment inherently uncertain prior hazard and loss models with an integrated strategy for updating these priors with ground-truth observations, thereby greatly reducing their uncertainties. Two strategies are employed. Early reports of casualties are used in a Bayesian updating fashion to constrain the possible range of fatalities and to lower the prior models' uncertainties. Additionally, remotely sensed satellite radar data, in the form of a Damage Proxy Map (or DPM), are used in a Bayesian causal graph framework combined with machine learning to optimize the mapping among the physical processes that cause shaking-based building damage, landslides, and liquefaction to prior expectation models. The casual graph framework also affords the potential for removing anthropogenetic noise contained in the imagery. Ultimately, our two-fold model updating strategy will accommodate key ground-truth observations such as fatality reports, locations of building damage, and ground failure reports to converge on actual losses more rapidly.

Conference Paper↗

USGS2021

Bayesian updating of seismic ground failure estimates via causal graphical models and satellite imagery

Earthquake-induced secondary ground failure hazards, such as liquefaction and landslides, result in catastrophic building and infrastructure damage as well as human fatalities. To facilitate emergency responses and mitigate losses, the U.S. Geological Survey provides a rapid hazard estimation system for earthquake-triggered landslides and liquefaction using geospatial susceptibility proxies and ShakeMap ground motion estimates. However, the resolution and accuracy of these models are often limited by coarse-granularity and large uncertainties of available geospatial features provided at a regional scale. Recently, with the advancement of remote sensing technologies, synthetic aperture radar (SAR) images are captured and analyzed to obtain a rapid estimate of earthquake-induced correlation changes between pre- and post-event images. These correlation changes indicate ground failures and building damage t, showing the potential to provide supplementary information for rapid hazard and loss estimation. However, the exact causes of changes in satellite images are not directly ascertained by the DPM alone. For example, changes could be due to building damage, landslides, liquefaction, noise or any combination thereof. More importantly, the occurrence and intensity of landslides, liquefaction, and building damages are spatially correlated, which makes it yet more challenging to distinguish the sources of any such changes. In this study, we develop a generalized causal graph-based Bayesian Network that models the physical interdependencies between geospatial features, seismic ground failures and building damage, as well as DPMs. Geospatial features provide physical insights for estimating ground failure occurrence while DPMs contain event-specific surface change observations. This physics-informed causal graph incorporate these variables with complex physical relationships in one holistic Bayesian updating scheme to effectively fuse information from both geospatial models and remote sensing data. This framework is scalable and flexible enough to deal with highly complex multi-hazard combinations. We then develop a stochastic variational inference algorithm to jointly update the intractable posterior probabilities of unobserved landslides, liquefaction, and building damage at different locations efficiently. In addition, a local graphical model pruning algorithm is presented to reduce the computational cost of large-scale seismic ground failure estimation. We apply this framework to September 2018 Hokkaido Iburi-Tobu, Japan (M6.6) earthquake and January 2020 Southwest Puerto Rico (M6.4) earthquake to evaluate the performance of our algorithm

Conference Paper↗